Distributed power generation is a concept that has been the subject of much discussion over the years, but to date there has not been widespread deployment of distributed generation systems. Distributed generation refers to the use of small electrical power generation systems located at the sites where the power is needed, and thus is to be distinguished from the traditional utility grid system in which a large central power plant produces power that is then transmitted over substantial distances to a plurality of users through a system of power transmission lines commonly referred to as a grid. In contrast to conventional power plants operated by utilities, which often can produce several megawatts of power, distributed generation systems are generally sized below two megawatts, and more typically in the 60 to 600 kilowatt range.
The failure of distributed generation to achieve widespread deployment can be attributed primarily to cost. In most areas of the United States, and indeed in much of the world, it simply has been cheaper for most users to purchase power from the grid than to invest in and operate a distributed generation system. A major factor driving the relatively high cost of power from distributed generation systems has been the relatively low efficiency of the small engines used in such systems, particularly at part-load operation conditions.
Typically, the generator in a distributed generation system is driven by a small turbine engine, often referred to as a microturbine or miniturbine depending on size. A turbine engine generally comprises a combustor for burning a fuel and air mixture to produce hot gases by converting the chemical energy of the fuel into heat energy, a turbine that expands the hot gases to rotate a shaft on which the turbine is mounted, and a compressor mounted on or coupled with the shaft and operable to compress the air that is supplied to the combustor. Because of the relatively small amount of electrical power required from a distributed generation system, the turbine engine is correspondingly small. For reasons relating to the aerodynamics that take place within the engine, and other reasons, the efficiency of a turbine engine tends to decrease with decreasing engine size. Accordingly, microturbines and miniturbines automatically have an efficiency disadvantage relative to larger engines.
Furthermore, irrespective of size, the part-load efficiency of a turbine engine is notoriously poor, generally because of the particular manner in which the engine is operated at part-load conditions. More particularly, it is typically the case in turbine engines that the turbine inlet temperature, which essentially represents the peak temperature of the working fluid in the engine cycle, falls as the power output from the engine decreases below the “design” point. The design point is typically a 100 percent rated load condition, and the engine is usually designed so that its peak efficiency occurs substantially at the design point. It is well known that the primary variable influencing the efficiency of the thermodynamic cycle of an engine is the peak temperature of the working fluid. All other things being equal, the greater the peak temperature, the greater the efficiency; conversely, the lower the peak temperature, the lower the efficiency. Therefore, if the engine, when operating at a part-load condition, is controlled in such a manner that the peak effective temperature of the working fluid in the cycle (i.e., the turbine inlet temperature) is substantially lower than what it is at the design point, the efficiency of the engine tends to suffer to a substantial extent.
In some prior-art gas turbines, particularly aircraft gas turbine engines for propulsion and large gas turbines for constant-speed electrical generator systems, variable-geometry systems have been used at part-load conditions to reduce the air flow rate so that engine efficiency does not unduly suffer. For instance, variable inlet guide vanes (IGVs) have been used in axial-flow compressors; at part-load conditions, the IGVs are closed down to reduce the air flow rate for a given compressor speed. In the case of radial compressors, the stator vanes have sometimes been made variable for achieving a similar effect. In still other cases, variable first-stage turbine vanes or nozzles have been employed for controlling the speed of the turbine and, hence, the speed of the compressor, so as to control air flow rate. Such variable-geometry systems are expensive, and the bearings and other movable components are prone to wear, thus making these systems impractical for electrical generation systems that must be available for service for a high percentage of hours per year, must be able to operate essentially continuously if required, and must also be able to respond quickly to changes in the power required by the load being served. Furthermore, variable-geometry mechanisms are not practical to implement in microturbines and miniturbines because of the small size of the engine. Thus, a need exists for an alternative to variable-geometry methods for optimizing engine performance at part-load conditions.
Emissions (including but not limited to nitrogen oxides, unburned hydrocarbons, and carbon monoxide) represent another aspect of distributed generation that has proven challenging. In general, for a given power output, NOx emissions tend to be reduced or minimized by minimizing the temperature of combustion of the fuel (also known as the flame temperature), which in general is higher than the peak thermodynamic temperature (turbine inlet temperature), thus reducing the production of oxides of nitrogen without adversely affecting efficiency. The primary method of reducing the flame temperature is to premix the fuel and air prior to the combustion zone to produce a mixture with a high relative ratio of fuel to air, i.e., a lean mixture. The premixing also assures that the temperature throughout the flame zone is very nearly uniform without hot spots that can lead to the local production of NOx. However, as the mixture is made leaner, carbon monoxide (CO), unburned hydrocarbon (UHC), and pressure fluctuations increase. These trends continue and the flame zone becomes more unstable as the mixture is made still leaner, until the lean extinction limit is reached. For mixtures any leaner than this limit, no flame can be sustained. In practice, carbon monoxide and unburned hydrocarbon emissions and/or pressure pulsations become unacceptably high before the lean extinction limit is reached.
The lean extinction limit may be moved to leaner regimes by increasing the inlet temperature to the combustor and by using catalytic combustion. The use of catalytic combustion substantially increases the operating regime for lean premixed combustion, resulting in very low NOx emission, acceptable CO and UHC emissions, and essentially no pressure pulsations. Catalytic combustion does, however, introduce another constraint on operation called the lower catalytic activity limit. The inlet temperature to the catalytic combustor must be kept above this limit to sustain catalytic combustion.
In many conventional microturbines, the engine control is such that at part-load conditions the combustor inlet temperature tends to fall and the fuel/air mixture becomes leaner. In the case of conventional lean pre-mixed combustion, this tends to result in increased emissions; in the case of catalytic combustion, the falling combustor inlet temperature can lead to failure to sustain catalytic combustion. In practice, lean-premixed and catalytic combustors are able to operate over only a portion of the load range of the gas turbine because of falling combustor inlet temperatures and the progressively leaner conditions that prevail as load is decreased.
In some cases, pre-burners have been used before combustors for boosting the combustor inlet temperature. Additionally, variable-geometry combustors have been used in which a portion of the air is diverted around the combustor to maintain the fuel/air ratio at a level allowing operational stability. The pre-burner solution poses a reliability penalty in that over-temperature or other malfunction of the pre-burner can damage the main burner, and also adds to the cost of the system. In addition, it imposes an operating cost penalty as a result of the pressure loss that occurs through the pre-burner; this pressure loss is experienced even when the pre-burner is not in use. Variable geometry can be applied to eliminate the pressure loss penalty in addition to its use in maintaining fuel/air ratio. However, variable geometry solutions are costly, complicated, and prone to excessive wear, decreasing reliability and increasing maintenance costs.
For many potential users, these factors have combined to make electrical power generation via distributed generation systems less attractive than purchasing power from the large utilities.